分析課題組文章03-xx

1、-GDIEngineOperationwithAqueousnol/GasolineBlendsand ON,FelixLEACH,FanRichardSTONE*,LongfeiCHEN,n (DepartmentofEngineeringScience,UniversityofOxford, Oxford, OX1 3PJ, Abstract: nolisbeingpromotedasarenewablefuelandasameans ofimprovingenergysecurity. AE85 (85%noland15%unleadedgasolineby volume) isnoww

2、ellestablishedinUSAandHowevertheremovalofwaterfrom nolis ensive,so-GDIEngineOperationwithAqueousnol/GasolineBlendsand ON,FelixLEACH,FanRichardSTONE*,LongfeiCHEN,n (DepartmentofEngineeringScience,UniversityofOxford, Oxford, OX1 3PJ, Abstract: nolisbeingpromotedasarenewablefuelandasameans ofimprovinge

3、nergysecurity. AE85 (85%noland15%unleadedgasolineby volume) isnowwellestablishedinUSAandHowevertheremovalofwaterfrom nolis ensive,so it isusefultoknowmoreabouttheperformanceofdifferentwaternolblends,andthemiscibilityofwaterwithnol/gasolineblends. authorstestedtheblendsofgasoline andnolfrom0100%nol,a

4、ndevaluatedtheir characteristics,combustionperformance,andparticulates. Inadditionblends ofnolandwater upto40%byvolumewater havebeentested. Ternaryphasediagramsasoline,nol,andwaterbeenused. Theresultstpresenceof waterinnol/gasolinemixturesisnot sarilyimpedimenttotheiruseasafuelinautomotiveengines. A

5、ddingnoltogasoline increasedthevolumeoftwasinjected andthepersistenceofthefuelsprays, lly foracoldengine. Thisledtoreducedhomogeneityandahecombustionstability,andanincreaseinparticulatemattersfor stoichiometricmixture. Addingwatertonolfurtherincreasedthevolumeoftwasinjected,butincreaseincombustiondu

6、rationandreductionincombustionstabilitywerenotsignificant withupto30%byvolume. forKeywords: gasolinedirectinjection（GDI）engine; particulateemis sprays combustions; blends; nol; aqueous RichardSTONE*,LongfeiGDI, nON,FelixLEACH, FanXU/ n ASa Journal,Weneed(OX1 3PJ, 能以及水與乙醇的溶混該文作者試驗了含乙醇 0100%、乙醇混合物，評價了

7、混合物的噴霧特性。向噴：收稿日期Received: 2012-06-基金項目Supportedby: theEngineeringandPhysical第一作者 ResearchCouncilandJaguarLandRover : HYPERLINK mailto:Richard.stoneeng.ox.ac.uk :第二作者Secondauthor：LongfeiCHEN。,講，Enolhasbeenusedextensivelyinautomotivefuelsprimarilyasa fuelextenderandanti-agentinnolgasolineblends,withnu

8、merousadvantagesincludinghigheroctane(allowinghighers,aswellasratios),reductionsinunburnedhydrocarbon(HC),COanderallairpollutionandde andonpetroleumfuels,e.g.Al-BaghdadiPurenolfuel(E100)hasalsobeenstudiede.g.Brewster2,de onstratinghigherspecificandbrakeefficiency,aswellaslowerCO2 ionofsugars,Enolism

9、ostlysimplymadefromtheompetitionEnolhasbeenusedextensivelyinautomotivefuelsprimarilyasa fuelextenderandanti-agentinnolgasolineblends,withnumerousadvantagesincludinghigheroctane(allowinghighers,aswellasratios),reductionsinunburnedhydrocarbon(HC),COanderallairpollutionandde andonpetroleumfuels,e.g.Al-

10、BaghdadiPurenolfuel(E100)hasalsobeenstudiede.g.Brewster2,de onstratinghigherspecificandbrakeefficiency,aswellaslowerCO2 ionofsugars,Enolismostlysimplymadefromtheompetitionwithfoodtsecondgenerationorcellulosic-nolneedstobeexploited. nolhasproducedfromcelluloseforover100years,butthereisnowalisationoft

11、hes. Lignocellulosicmaterials(suchaswood)havetobefinelyandthenhydrolysed(byanacid) toreleasethecellulose. Enzymesthenbreakthesimplechasglucose,whichcanbetoproducenol. Ov rthe cours manymillenniayeastshavebeenselectivelytenableahighconcentration(overbyvolume)ofnoltobeproducedbyioninaqueousThenolcanbe

12、partiallyseparatedbydistillation,butwaterandnolformanmixture. Anazeotropicmixtureisoneinwhichtheitionoftheliquidandvapouridentical,sothislimitstheseparationbydistillation. Theazeotropehasaitionenolbyvolume(95.6%bymass)andiscalledhydrousnol. Waterboilsat100C, boilsat78.4C,buttheazeotropeboilsat78.2 C

13、,whichisneitherofits constituentsthisisknownasitiveazeotrope). Theproductionofanhydrousnolisbytwolythereisazeotropicdistillation,whichdependsonaddingbenzeneorcyclohexanethemixture;theseareselectedsincetheywillformaternaryazeotrope. Withcyclohexaneasentrainer,justsufficientcyclohexaneisaddedtoincorpo

14、ratealltheothe azeotrope(7%water,17%nol,and76%cyclohexane),whichboilsat62.1C. Whenmixtureisdistilleditanhydroushecolumnbottom(butwithcontaminationbythebenzene orcyclohexane),andavapourmixtureofwatercyclohexane/benzene,whichcanbe condensedtoformatwo-phaseliquidmixturefromwhichiteasytoseparatethewater

15、. Thesecondmorerecent methodismuchlessensiveandjustlikegaseousreswingabsorptionitusesmolecularsievestoremovewaterfromeEnol. Thenolvapourhemolecularsieveallowabsorptionofwaterwhiletheyexcluderiodreispassedthroughabedofmolecularsievebeads. time,thebedisregeneratedbytheflowofmosphere(e.g.N2)orundervacu

16、umtotheabsorbedwater. Byobeds oneisavailabletoabsorbwaterwhiletheotherisHydrousnolcanbeusedasafuelalone,butunlikeanhydrousnol,hydrousnolismiscibleinallratioswithgasoline,sothewaterfractionistypicallyremoved priortouseinengines. AnnolblendknownasE85(meaning85 percentbyvolumenol)is availableinUSA,Swed

17、enandFinland;inBrazilpurenolisused(E100). Toestartinginerthenolcontentisreduced (E70inUSA,andE75inSweden),butitismarketedasE85. However,aswillbehenextsectionthereisscopeforusingenolandgasoline1Aqueousnoland GasolineWaterandnolarefullymiscibleasarenolandgasoline, butwaterandgasolineessentiallyimmisci

18、ble,sothisraisesthequestionofppenswhenwaterisadded toenolgasolinemixturesuchasE85,whatwouldhappenifaqueousnoltshowsblendedwithgasoline? Theanswerisprovidedbyternarychas Fig. region(atthetopofthefigure)whereasinglephaseisformed. Belowthephaseboundarymixturewillcomprisetwoseparatephases. Aswellasdepen

19、dingontemperature,itionitionof thegasoline. OwenandColey 3reportthephaseboundarydependsonthettincreasingthearomaticcontentincreasesthewatertoleranceofgasoline,co-solvents. JohansenandSchramm4arehigheralcohols(suchasiso-butanol)concernedwithhow toformtstayhesinglephaseregion. Theyt hesinglephaseregio

20、n,thenalineodifferentternaryscombinationofthesemixtureswillremainasasinglehelimitingcasethelinewilltangentialtothephaseseparation curve. UnlikeJohansenandSchramm4,the happens with phase separation.eresthereisFig. concernedwithhow toformtstayhesinglephaseregion. Theyt hesinglephaseregion,thenalineodi

21、fferentternaryscombinationofthesemixtureswillremainasasinglehelimitingcasethelinewilltangentialtothephaseseparation curve. UnlikeJohansenandSchramm4,the happens with phase separation.eresthereisFig. Ternarydiagramfornol-gasoline-waterblendtie-line (thebrokenline)toshowitionwhenamixtureformstwophases

22、; E73,W15,G12means73% nol,15%water,gasoline; adaptedfromJohansenandSchrammOnthephasew=fw(g)ande= fe(g)where w,eandgaretherespectivepercentageswater,nolandgasoline. e thereisamixtureof ition tis definedas andifthisfallshetwo-phaseregion,thentheitionsofthetwo phasesare(w1,g1,and(w2,g2,e2);letxbethefra

23、ctionofphase ByconservationoftheindividualG=xg1 +(1W=xfw(g1)+(1x)E=xfe(g1)+(1x)AsW, G,Eareknown(asarethe(g)andfe(g),thenthese3simultaneousequationsbesolvedtofindthe3unknowns,namelyx,g1,andg2,whichcanthenbeused tofindw1,e1 w2,Inotherwords,iftheoftheboundarynthesinglephaseandtwophaseregiontformstwopha

24、ses,thenthereisonlyknown,thenforamixtureofknown siblecombinationofitionsforthe twoseparatephases. Equilibriumwillrequireenergyofthetwophasestobeequal,andthisenablesitionofthe phaseboundary beConsiderE85(designatedastheyellowE85,W0,G15inFig.1,meaning85%nol,0%and15%gasoline)towhichwaterisadded. Aswate

25、risaddedthe itionfollowsastraighttterminatesatthe100%x. However,onlyacertainamountofwaterisenolgasolinemixturebeforethemixtureo2phases anolcontainingatyofgasolineandanolphasecontainingatywater. If17.5kgofwaterisadded to100kgof theE85(thedensitiesofnolandgasolinecanconsideredtobeequal),amixtureisform

26、ed(E73,W15,tisstillasinglephase,butifmorewaterisaddedthentwo phaseswillform. If50kgof waterisaddedto 100kgofE85then overallitionwouldbe(E57,W33,G10),butthisfallshe2phaseregion(2sinFig.1). TheiterativesolutionofEq.1andthe phaseboundarycoordinatest2Phase90%ofthemixturehasaition(E60, W37,G3),and10%ofth

27、emixturehasa (E30,W2,G68). Asacheckforeachof the3 Enol,0.9x60%+ 0.1x30% =Water,0.9x37%+ 0.1x2%= Gasoline,0.9x3%+0.1x 68%= Thelinearbehaviour90%ofthemixturehasaition(E60, W37,G3),and10%ofthemixturehasa (E30,W2,G68). Asacheckforeachof the3 Enol,0.9x60%+ 0.1x30% =Water,0.9x37%+ 0.1x2%= Gasoline,0.9x3%+

28、0.1x 68%= Thelinearbehaviourofternarydiagramsttheleverrules,thiscanitionof confirmedbyinspectionofthe tie-lineinFig.1(thebrokentdefinesthetwophases. Thiscanbereadilyseenbyconsidering traverseatie-taremadewithtExperimentalEquipment andFuelTheULGfuelhetests wasprovidedby.sabout22%aromaticswhichisntpre

29、sentinsomeEuropeanULGand sa correspondinglyercentageofiso-of64%. Italsohasa slightlylowerfinalboiling169C.Acomparisonoftheunleaded(ULG)andnolpropertiesisable1,and ofparticularsignificanceiswhenaremadeonaunitmassbasisforastoichiometricTablePropertiesofnolandDensity 15 C/(kg.m-Reid Vapor Prere / kPa R

30、esearch Octane Number Motor Octane Number BoilingPo/ ( C)-StoichiometricAirFuellpyofCombustion(Liq.)/(MJ.kg-1) Heat of Vap. / (kJ.kg-1)StoichiometriclpyofCombustion(Liq.)/(MJ.kg HeatofVap./(kJ.kg-Andersonetal.5haveprovidedacomprehensivereviewoftheoctaneratingdatafornolenol,andhighlightthereasonsfort

31、hehedata (inpartdue to procedures). Theyhavealsocarefullyreviewedtheoctaneratingdataforalcohol/gasolineand the use of blending octane numbers (bON, these are used to account for the non-linear mixing behaviour when alcohols and gasoline are blended). UnfortunaythevalueofthebONdependsontisbeingadded.

32、 Andersonetal. 5 thevolumefractionoftsincethemechanismswilldependonthepartialresofthehevaporisedfuel,thenoctaneratingofalcohol/gasolineblendsshouldbecalculatedbyusingamolarweightedThisremarkablysimpleprocedurethentthe octaneratingcohol/gasolinecanbecalculatedusingthe purealcoholRONandMONvalues.Itiss

33、arytoknowthe andmeanmolarmassofthegasoline,butAndersonetal. 5 onlyneededtousea 750 kg/m3 and 110 kg/kmol.medvaluesTheeffectonthevolatilityofaddingnoltogasolineiscomplexbecausethegasolineispolarandthenolispolarinsimpletermstheydonotwantto mix. So, despitethelowVaporre(RVP)of nol,whenitisadded to gaso

34、linetheRVPof themixturewill initiallyHsiehetal.6examinedtheRVPofnol/ULGblends. Withincreasingnolcontenttheinitiallyincreasestoumat10%noladdition,butthen EngineandTestItiswell tGasolineDirectInjectionenginescanoperatehighernPortFuelInjection(PFI)enginesbecausetheevaporativecoolingfromthefuelcanbeused

35、 coolingHsiehetal.6examinedtheRVPofnol/ULGblends. Withincreasingnolcontenttheinitiallyincreasestoumat10%noladdition,butthen EngineandTestItiswell tGasolineDirectInjectionenginescanoperatehighernPortFuelInjection(PFI)enginesbecausetheevaporativecoolingfromthefuelcanbeused coolingair. Sincethetemperat

36、ureisloweratthestartof ,thensubsequentseswillbeatalowertemperature,soforagivenoctaneratingfuel,thenaratiocanbeused foragivenchargetemperatureof . Thechargetemperaturealsoincreasesthevolumetricefficiencyleadingtoanheoutputandaconsequentialadditionalreductioninfuelconsumption. Theincreasedlpyvaporisat

37、ionwithnolblendscouldleadtoadditionalgains,butmixturesmightalsobehomogeneous. ThecombustionsystemisessentiallythesametheJaguarsbeenco prehensivelydescribedbySandfordetal.7.Theinjectionrewasbarandtheinjectiontiming（CA，bTDC）was280throughouttheTableSingle-CylinderGDIOpticalEngineCombustion FuelPre Igni

38、tion,CA(bTDC) SGDI, 4ValvesperMulti-holeNozzle 8990.3 mm15.0kPa 35 1 500 ManifoldAbsolute50 AseriesofinjectionandcombustionimageswascapturedbyaI-1024highcameraatup to9,000fps.IlluminationofthespraywassynchronizedLEDsystem,whilst swasvisualizedwiththevisiblelightflamechemiluminescence. singwasusedtos

39、eparatethe sprayfromthebackgroundandxelcountheegratedtogiveaivevaluefortheMieACambustionHFR400fFID(fastresponseFlameIonisationDetector)wasusedtoinvestigatevariabilityofthepre-flamehydrocarbonconcentrationnearthesparkplug,usinganoffsetplugthreadedholemodatealingcapillary. ParticulateMatter(PM)concent

40、rationsandsizedistributionsweremeasuredCambustionDMS500mobilityparticlesizer,usingtheCambustionlingsystemwhichincorporatesaheatedpipeanddilutionair. Thedilutedexhaustaerosolpassesthroughacorona-wireuni-polarroducingaknownchargedistributionontheaerosolparticles. Theseparticlesthenwhensubjectedtoanele

41、ctricfieldandimpactuponelectrometerringscontainedDMS500classifiercolumn.Thecurrentfromtheseringsisthenmeasuredandisproportionaltonumberofparticlesimpactingupontherings8. ThetransferfunctionoftheDMS500thentheseringoasizeresolvednumberconcentration. TheDMS500softwareusesBayesianisticalalgorithmtogener

42、atetwolognormaldistributionfittingspectra,whichmeanttorepresentthenucleationmodeandtheaccumulationmode.Thistechniquehasadditionaltittends toremovenoise. Thisisparticularlyimportantforlargesizes,forwhichevensmallnumberconcentrationsleadtolargemassvalues. PMconcentrationswerederivedfromPMnumberconcent

43、rationsusingaPMdensity:M(kg)=1.7210-24ThisequationhadbeenexperimentallydeterminedusingaCouetteCPMA(CentrifugalParticleysimilarSGDIcombustionyser)3 Resultsand3.1 nolandGasolineFigure2tfor the coldcondition,theincreasednolcontentproducedMsermsofboth totalnumberandtotalmass;thetrendfor thewarmcondition

44、isclear. The3 Resultsand3.1 nolandGasolineFigure2tfor the coldcondition,theincreasednolcontentproducedMsermsofboth totalnumberandtotalmass;thetrendfor thewarmconditionisclear. ThemainreasonfortheseabnormalitiestULGanditsE10blendproduce verylowofPMsn1orderofmagnitudenbackgroundnoiselevel)duetothe aro

45、maticitionandhence thehighnoisetosignalratioveconfoundedtheresults. Also,anyslightdifferenceinAFRwouldaffecttheresults. Insummary,foramixture,E856-15timeshighertotalPMnumberand8-11 timeshighertotalPMnULG.ThetotalPMmasswiththeE85blendis11twithULGunderthe coldconditionitis8timeshigherunderthewarmcondi

46、tion. ThesephenomenaareprimarilyattributedtohefuelsprayandsubsequentvaporisationDifferent(a) TotalParticulate(b) TotalParticulateTotalparticulatenumber(Pn,top)andtotalparticulatemass(Pm,bottom)emis stoichiometricnolblendsinacold(20C )andawarm(80C)Fig. sfor Figure3longer,tas thenol hegasolineisincrea

47、sed,thenthesprayspersistllyforcoldengineoperation. Thenolincreasesthevolumeofs toinjectedandthelpyofvaporisation. Itcanbetthiswillincreaseinhomogeneityofthemixture,causingareductionincombustionstabilityandthePMs(a) Cold Condition(At 20 (b) WarmCondition(At 80(a) Cold Condition(At 20 (b) WarmConditio

48、n(At 80 Bars tedwiththe averagevaluesare1standardFig. egratedpixelvaluesasafunctionofangleforthecoldandwarmconditionforstoichiometricnolDifferentFig. CoVoftheIMEPwithdifferentstoichiometricULG/e nolblendsforthecold(20C )andwarm(80)Thecolorofbars should besameasFig. 2. Table3duration,andthighlevelsof

49、nolhecoldengine)increasedtheurn(seeFigure4)wastedreducedcombustionTable3 Ensembleaveragevaluesofstoichiometriccombustionparametersforthedifferent nolinacold(20C)andwarm(80)engine;ignitionatTable3duration,andthighlevelsofnolhecoldengine)increasedtheurn(seeFigure4)wastedreducedcombustionTable3 Ensembl

50、eaveragevaluesofstoichiometriccombustionparametersforthedifferent nolinacold(20C)andwarm(80)engine;ignitionatCA /CA /IMEPIMEP0-10%10%-80%0-10%10%-80%MFBmassfractionFigure5showstypicalsignalsfromthein-cylinderfFIDlingheregionoftheplug.ThefFIDsignalriseslyduringthestroke,butthenrisesmoreslowlycombusti

51、onoccurs,atwhichthehydrocarbonlevelfallsly. umvalueoffFIDsignalwasusedtorepresentthehydrocarbonlevel,andrfectlyhomogeneousthenthesevaluesshouldbeconstantas theoverallairfuelratiowasThecycle-to-fiedbyusinghefFIDsignalthusindicatesinhomogeneity,andthisCoefficientofvariation(CoV=standardFig. An leofpre

52、-flamefFIDsignals)andthe in-re)stoichiometricULGat80C ;notransportdelaycorrectionhasbeenapp dtothefFIDFig. 6thighlevelsofnolhecoldengine)increasedthehecylinder,andthisappearstocorrelatewellwiththeincreasedPMs(Fig. Inrealworldvehicletestsitiscommonforvehiclesfuelledonnolblendstoproduceraisheretal.10.

53、 Oneexplanationforthismassandnumbers,seeforttherewill beixturetransientsduringacceleration(eithererissought,orbecauseofthedifficultyincontrollingtheAFRduringtransients). TheeffectofAFRPMsable4,forwhichthe basefuelwasa65%iso-octane/35% Fig. CoVofthepre-flamefFIDsignalsforthedifferent stoichiometric U

54、LG/e (20 C ) and warm (80 C ) engine.nolfuelblends inaThecolorofbars should besameasFig. 2. TableTheeffectofstoichiometryonthetotalparticulatenumber andparticulatemass(ignition35 bTDC); the base fuel is 65% iso-octane/35% tolueneTotalTotalMassFig. CoVofthepre-flamefFIDsignalsforthedifferent stoichio

55、metric ULG/e (20 C ) and warm (80 C ) engine.nolfuelblends inaThecolorofbars should besameasFig. 2. TableTheeffectofstoichiometryonthetotalparticulatenumber andparticulatemass(ignition35 bTDC); the base fuel is 65% iso-octane/35% tolueneTotalTotalMasstforless airfuel ratioTable4ixtureoperationtheE10

56、fuelhadmassstwereaboutorderofmagnitudenthebasefuel,whilethenumberswithE10(potentiallygreaterchallengeifafter-treatmentistobeavoided) werenearly2ordersofmagnitudenthebasefuel. Thistanyixtureoperationinadrivecyclewillyresponsibleforparticulates,tanything(such asnoltparticulatesduringixtureoperationwil

57、lleadtoanoverallreductionparticulates,evenifthestoichiometricsare2.2 nolandWaterBlendswithupto40%byvolumewaterhavebeentested,asW0(E100),W10,W20,W30andalongwithE85.Theirperformancewasscattering),combustionperformance(fromin-ermsofthefuelspray(measuredbyreysis)andmatter(PM)s. Thedifferenttestcondition

58、swereasTypeofInjection: GasolineDirectInjection(GDI)orPlenumFuelInjectionLambdalessAirFuelRatio(AFR):0.9orFigure7showstheegratedpixelvaluesfromtheMiescatteringimagesfornol(E100), nol(W30)andE85nolLambdalessAirFuelRatio(AFR):0.9orFigure7showstheegratedpixelvaluesfromtheMiescatteringimagesfornol(E100)

59、, nol(W30)andE85noland15%gasoline);thebarsrepresent 1deviation.TheMiescatteringimagesshowtheevolutionandofthefuelspraysasafunctionly,whiletheW30hasboththe lpy of vaporization 2.4 timesangle,andaswouldbeexpectedtheE85dispersesthemosttyofmixturetobeinjectedandthehighestspecifict oftheFig. Theegratedpi

60、xelvaluesoftheMiescatteringtshowtheevolutionandthefuelspraysasafunctionof anglefornol(E100),30%water/70%nol(W30)andnoland15%gasoline)plenuminlettemperatureof30 C;thebarsrepresent standard 1)Thenumbers notedatthe horizontalaxis shouldbelike280, 270,260,tmeanscarry-overevery510, Not every3.2) Threelin